AIbase
EN
Home
Swin Tiny Patch4 Window7 224 Finetuned New Dataset 50e
S

Swin Tiny Patch4 Window7 224 Finetuned New Dataset 50e

Developed by Gokulapriyan
Image classification model based on Swin Transformer Tiny architecture, fine-tuned for 50 epochs on a custom dataset with 79.73% accuracy
Image Classification
Transformers
Open Source License:Apache-2.0#Image Classification Fine-tuning#Swin-Transformer Architecture#High Accuracy Model
Downloads 17
Release Time : 2/8/2023

Model Overview

This model is a vision Transformer based on Microsoft's Swin Transformer Tiny architecture, specifically optimized for image classification tasks. Through 50 epochs of fine-tuning on a specific dataset, the model demonstrates strong image classification capabilities.

Model Features

Efficient Transformer Architecture
Utilizes Swin Transformer's hierarchical window attention mechanism to maintain high performance while reducing computational complexity
Transfer Learning Optimization
Fine-tuned from a pre-trained model, achieving good results even on small-scale datasets
Balanced Performance and Efficiency
The Tiny version strikes a good balance between computational resources and model performance, suitable for practical deployment

Model Capabilities

Image Classification
Visual Feature Extraction
Transfer Learning

Use Cases

Computer Vision
General Image Classification
Classify and recognize various types of images
Achieved 79.73% accuracy on the test set
Domain-Specific Classification
Can be fine-tuned for specific domains (e.g., medical imaging, industrial inspection)
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
EnglishįŽ€äŊ“中文įšéĢ”中文ãĢãģんご
Š 2025AIbase